Department of Transportation

Settlement Analysis and Mine Subsidence

Settlement of our roads, bridges, and other structures is to be minimized through realistic estimates and forethought in design. Embankment settlement is problematic in Pennsylvanian aged red beds and soft shales that tend to break down over time. Approach slab settlement can create a “bump-at-the-end-of-the-bridge.” Bridges in West Virginia are normally supported by bedrock and, for most of our bridge sizes, the elastic settlement of the bedrock is negligible; however, for our high bridges, the tilt of piers can present significant side-sway. Differential settlement and distortion to structures are to be considered. Settlement induced downdrag and bending of battered piles can be a concern. Mine subsidence does not fit into the normal geotechnical analysis, but it does occur in West Virginia and requires a deliberate decision as to the risk and cost to remediate subsidence.

Settlement of embankment rock fills is common up to a rate of 0.005 feet/foot when properly placed and compacted. Proper placement of rock benches, blankets, and underdrains can prevent water from building up and causing the subsequent secondary long-term settlement of the fill material through breakdown of soft rocks. Special consideration is needed by the geotechnical engineer where temporary haul roads are placed in rough terrain to gain access into lower fill areas, and especially in cut/fill transition areas. These temporary fill areas, that are within the embankment foot print, need to be removed and compacted in controlled lifts per the WVDOH Standard Specifications. Where soft shale and claystone is encountered, these materials require segregation and placement in thin lifts as random fill. Fills to be placed on soft/wet foundation soils should be placed at a staged rate (6 vertical feet/week is recommended) to avoid overloading the soil. Alternatively, a contractor and his geotechnical engineer can use piezometers to monitor the pore pressures to control the rate of fill placement. Wick drains that are properly designed and constructed can be used to accelerate the rate of settlement. Quarantine times before either pile, or pavement placement can be an effective settlement mitigation; provided the construction schedule permits it. Alternatively, a contractor can monitor settlement of fills by precision surveying to verify that the rate of settlement is acceptable (usually 0.01 feet / 2 weeks). Differential settlement of drainage structures and utilities is to be considered in the design. Differential settlement of a reinforced soil slope (RSS) needs to be checked to ensure that the maximum strain in the reinforcement layers is not exceeded. The geotechnical engineer is to ensure his recommendations are noted in the final plans.

Approach slab differential settlement is to be minimized by proper placement and compaction of approach fill. To ensure drivers’ comfort, differential settlement across a 20-foot long slab should be kept to 1 inch or less over the life of the bridge. If the estimated settlement exceeds this criteria, mitigation of settlement is needed. Reinforced backfill can minimize settlement of the fill immediately beneath the slab, but is not a substitute for proper compaction of the larger embankment fill nor for addressing soft foundation soils in design.

Downdrag induced loading to deep foundations is to be considered when settlement occurs and is estimated as negative skin friction. Downdrag loading is to be factored according the method used to estimate it and whether it is being used to either resist uplift or added to the loading. Downdrag is to be considered over the surface area along the depth of the foundation where the displacement is greater than 0.4 inches.

Differential settlement of bridge structures supported by bedrock is normally insignificant for our smaller bridges. However, for large multi-span bridges with significantly different rock moduli at different substructures, significant differential elastic settlement and the resulting load effects can be present, especially during construction. The commentary in LRFD C10.5.2.2 limits differential settlement and the resulting angular distortion to 0.008 feet/foot and 0.004 feet/foot for simple and continuous span bridges, respectively. This angular distortion is evaluated at the top of the substructures.

High piers that are laterally loaded by live loads can induce a triangular stress distribution within the bedrock which can result is differential settlement that is multiplied over the height of the pier and can add significantly to the bridge side-sway. Refer to LRFD for the equations for elastic settlement of rock. The geotechnical engineer is to estimate the rock mass modulus using the RMR method instead of the GSI method, when needed. The lateral movement is measured at the top of the substructure and is generally limited to 1.5 inches; however, the structural engineer can allow greater movement depending on the bridge/substructure(s) stiffness, bearing type, width, and thickness.

Differential settlement along MSE walls and other walls that are supported by soil should be estimated. The maximum distortion for the product as specified by the manufacture is not to be exceeded. As a maximum for MSE walls, the differential settlement should be no more than 1 foot per 250 feet.

Consolidation settlement of soft, wet, cohesive soils (N-value is 4 or less) supporting structures and pavement fill is problematic and should be estimated. The geotechnical engineer may waive the need to estimate consolidation settlement when fill heights do not exceed 10 feet. Preconsolidation pressure of the layer(s) in question is to be estimated to determine whether the soil is normally or overly consolidated and the compression index and/or recompression index is to be used accordingly. These indices and the coefficient of consolidation are to be determined by laboratory odometer testing or estimation from index properties. Plotted laboratory data is to be reviewed by the geotechnical engineer and only appropriate representative curves are to be use in the analysis. For some curves, graphical adjustment by standard methods may be needed to correct computer generated plots (e.g., Pc’, d100 and T90 as appropriate). Time rate estimates are to consider the length of the longest drainage path or paths when multiple drainage is present. To estimate the vertical stress intensity at mid-height in the layer in question, the Boussinesq graphical method is appropriate. For large embankment and approach fills, graphical methods such as presented in the FHWA NHI-06-088, Soils and Foundations, should be used to estimate the stress increase.

Settlement of sand, sand and gravel, and silt using corrected N-values and the Hough method can be used. Otherwise use the elastic method presented in LRFD.

Mine subsidence and its vertical and horizontal displacements are best understood through computer modeling. Where subsidence is likely, life-cycle cost analysis comparing the grouting costs to the costs of releveling and repairing the subsided bridge, in view of the probability of subsidence over the lifespan of the bridge, should be performed. If grouting is warranted, then an angle of draw between 15° and 30° is selected for the grounding zone based on the tolerable displacements of the structure. For single span bridges, it is usually more cost effective to decouple the superstructure from the substructure (semi-integral) and allow for future jacking and releveling of a structure instead of grouting when predicated vertical displacement is under 1 foot.

Foundation settlement and side-sway induced loads are to be factored as indicated in Table 3.4.1-1 of the LRFD.